Matematiska modeller för beslutsfattande hos människor, djur och maskiner
Åström, Kalle LU
(2020)
In Statsvetenskaplig tidskrift
122(4).
p.531-543
- Abstract
- The science of decision making is studied in many areas of science and engineering, for example within, computer science, automatic control theory, mathematics and neurophysiology. The aim of this text is to provide a very brief overview of some of the models and tools used within mathematical decision making. These models can be studied in terms of an agent that observes the world and takes decisions. The techniques that have been developed, e g optimal control, dual control, machine learning, reinforcement learning, have found numerous uses in many applications, where machines take decisions, but they are also an inspiration for the question of understanding biological decision making, and could potentially be important for... (More)
- The science of decision making is studied in many areas of science and engineering, for example within, computer science, automatic control theory, mathematics and neurophysiology. The aim of this text is to provide a very brief overview of some of the models and tools used within mathematical decision making. These models can be studied in terms of an agent that observes the world and takes decisions. The techniques that have been developed, e g optimal control, dual control, machine learning, reinforcement learning, have found numerous uses in many applications, where machines take decisions, but they are also an inspiration for the question of understanding biological decision making, and could potentially be important for understanding decision making in political science.InledningTypiskt för modellering och särskilt matematisk modellering är att man behö-ver göra förenklingar av de fenomen man vill betrakta. Det betyder ofta att en modell bara fångar vissa egenskaper hos det man vill efterlikna. En leksaks-modell av en bil kanske är användbar för att beskriva färg och form, men kan inte användas för att studera t ex bränsleförbrukning. Typiskt för modellerings-processen är att man, trots att man vet att modellen har brister, genom analys försöker förstå vilka slutsatser som har bäring på det väsentliga eller centrala problemet. Det är då viktigt att man försöker förstå vilka aspekter som är rele-vanta och vilka som är osäkra eller direkt irrelevanta.Inom maskininlärning pratar man t ex också om hur väl en modell generaliserar till nya data.
(Less) - Abstract (Swedish)
- The science of decision making is studied in many areas of science and engineering, for example within, computer science, automatic control theory, mathematics and neurophysiology. The aim of this text is to provide a very brief overview of some of the models and tools used within mathematical decision making. These models can be studied in terms of an agent that observes the world and takes decisions. The techniques that have been developed, e g optimal control, dual control, machine learning, reinforcement learning, have found numerous uses in many applications, where machines take decisions, but they are also an inspiration for the question of understanding biological decision making, and could potentially be important for understanding... (More)
- The science of decision making is studied in many areas of science and engineering, for example within, computer science, automatic control theory, mathematics and neurophysiology. The aim of this text is to provide a very brief overview of some of the models and tools used within mathematical decision making. These models can be studied in terms of an agent that observes the world and takes decisions. The techniques that have been developed, e g optimal control, dual control, machine learning, reinforcement learning, have found numerous uses in many applications, where machines take decisions, but they are also an inspiration for the question of understanding biological decision making, and could potentially be important for understanding decision making in political science. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/ddd619df-1956-4879-b4af-5d9344466353
- author
- Åström, Kalle
LU
- organization
- alternative title
- Mathematical models for decision-making in humans, animals and machines
- publishing date
- 2020-12-31
- type
- Contribution to journal
- publication status
- published
- subject
- in
- Statsvetenskaplig tidskrift
- volume
- 122
- issue
- 4
- pages
- 531 - 543
- publisher
- Fahlbeckska stiftelsen
- ISSN
- 0039-0747
- language
- Swedish
- LU publication?
- yes
- id
- ddd619df-1956-4879-b4af-5d9344466353
- alternative location
- https://journals.lub.lu.se/st/article/view/22354
- date added to LUP
- 2021-05-03 15:31:58
- date last changed
- 2022-01-29 02:36:51
@article{ddd619df-1956-4879-b4af-5d9344466353,
abstract = {{The science of decision making is studied in many areas of science and engineering, for example within, computer science, automatic control theory, mathematics and neurophysiology. The aim of this text is to provide a very brief overview of some of the models and tools used within mathematical decision making. These models can be studied in terms of an agent that observes the world and takes decisions. The techniques that have been developed, e g optimal control, dual control, machine learning, reinforcement learning, have found numerous uses in many applications, where machines take decisions, but they are also an inspiration for the question of understanding biological decision making, and could potentially be important for understanding decision making in political science.InledningTypiskt för modellering och särskilt matematisk modellering är att man behö-ver göra förenklingar av de fenomen man vill betrakta. Det betyder ofta att en modell bara fångar vissa egenskaper hos det man vill efterlikna. En leksaks-modell av en bil kanske är användbar för att beskriva färg och form, men kan inte användas för att studera t ex bränsleförbrukning. Typiskt för modellerings-processen är att man, trots att man vet att modellen har brister, genom analys försöker förstå vilka slutsatser som har bäring på det väsentliga eller centrala problemet. Det är då viktigt att man försöker förstå vilka aspekter som är rele-vanta och vilka som är osäkra eller direkt irrelevanta.Inom maskininlärning pratar man t ex också om hur väl en modell generaliserar till nya data.<br/>}},
author = {{Åström, Kalle}},
issn = {{0039-0747}},
language = {{swe}},
month = {{12}},
number = {{4}},
pages = {{531--543}},
publisher = {{Fahlbeckska stiftelsen}},
series = {{Statsvetenskaplig tidskrift}},
title = {{Matematiska modeller för beslutsfattande hos människor, djur och maskiner}},
url = {{https://journals.lub.lu.se/st/article/view/22354}},
volume = {{122}},
year = {{2020}},
}